基于直觉模糊聚类的目标识别  

Target recognition based on intuitionistic fuzzy clustering

作　　者：贺正洪[1] 雷英杰[1] 王刚[1] 

机构地区：[1]空军工程大学导弹学院,陕西三原713800

出　　处：《系统工程与电子技术》2011年第6期1283-1286,共4页Systems Engineering and Electronics

基　　金：国家自然科学基金(60773209);陕西省自然科学基金(2010JM8013)资助课题

摘　　要：空中目标的正确识别对防空作战具有重要的意义。针对目标识别的特点,提出了基于直觉模糊聚类的识别方法。首先,给出了直觉模糊相似度和相异度的度量方法,并利用目标特征信息计算直觉模糊相似度和相异度。然后,在分析了现有的几种直觉模糊相似矩阵构造方法的基础上,提出了一种新的直觉模糊相似矩阵的构造方法,并给出了直觉模糊相似矩阵转化为直觉模糊等价矩阵的途径;此外,定义了直觉模糊等价矩阵的(α,β)截矩阵,进而实现直觉模糊集的聚类。最后,运用该方法对给定特征的空中目标进行了聚类分析,实验结果表明,提出的方法是可行、有效的。The validity of air target recognition is very important for air defense combat.Concerned with the characteristics of air target recognition,intuitionistic fuzzy clustering is presented.Some valid measures of similarity and dissimilarity are griven;and with target characteristics,intuitionistic fuzzy similarity and dissimilarity are obtained.Serveral existing construction methods of the intuitionistic fuzzy similarity matrix are analyzed,a new construction method of intuitionistic fuzzy similarity matrix is put forward.An approach to transforming the intuitionistic fuzzy similarity matrices into the intuitionistic fuzzy equivalence matrices is given.Moreover,the（α,β） cutting matrix of intuitionistic fuzzy equivalence matrix is given,and a method of clustering intuitionistic fuzzy sets is proposed.Finally,based on this method,the clustering analyses of the air targets with specific characteristics are carried out.The result of experiment shows that this method is feasible and effective.

关 键 词：目标识别 直觉模糊集 直觉模糊聚类 直觉模糊相似矩阵 

分 类 号：TP182[自动化与计算机技术—控制理论与控制工程] TP391[自动化与计算机技术—控制科学与工程]